def test_pdf_reflected_for_negative_skewness(self):
sas_pos_skew = tfd.VectorSinhArcsinhDiag(
loc=[0.], scale_identity_multiplier=1., skewness=2., validate_args=True)
sas_neg_skew = tfd.VectorSinhArcsinhDiag(
loc=[0.],
scale_identity_multiplier=1.,
skewness=-2.,
validate_args=True)
x = np.linspace(-2, 2, num=5).astype(np.float32).reshape(5, 1)
self.assertAllClose(*self.evaluate(
[sas_pos_skew.prob(x), sas_neg_skew.prob(x[::-1])]))
def test_passing_in_laplace_plus_defaults_is_same_as_laplace(self):
d = 10
scale_diag = rng.rand(d) + np.float64(1.2)
loc = rng.randn(d)
vlap = tfd.VectorLaplaceDiag(loc=loc,
scale_diag=scale_diag,
validate_args=True)
sasvlap = tfd.VectorSinhArcsinhDiag(loc=loc,
scale_diag=scale_diag,
distribution=tfd.Laplace(
np.float64(0.),
np.float64(1.)),
validate_args=True)
x = rng.randn(5, d)
vlap_pdf, sasvlap_pdf = self.evaluate([vlap.prob(x), sasvlap.prob(x)])
self.assertAllClose(vlap_pdf, sasvlap_pdf)
vlap_samps, sasvlap_samps = self.evaluate([
vlap.sample(10000, seed=tfp_test_util.test_seed()),
sasvlap.sample(10000, seed=tfp_test_util.test_seed())
])
self.assertAllClose(loc, sasvlap_samps.mean(axis=0), atol=0.1)
self.assertAllClose(vlap_samps.mean(axis=0),
sasvlap_samps.mean(axis=0),
atol=0.1)
self.assertAllClose(vlap_samps.std(axis=0),
sasvlap_samps.std(axis=0),
atol=0.1)
def test_default_is_same_as_normal(self):
d = 10
scale_diag = rng.rand(d) + np.float64(1.0)
loc = rng.randn(d)
norm = tfd.MultivariateNormalDiag(loc=loc,
scale_diag=scale_diag,
validate_args=True)
sasnorm = tfd.VectorSinhArcsinhDiag(loc=loc,
scale_diag=scale_diag,
validate_args=True)
x = rng.randn(5, d)
norm_pdf, sasnorm_pdf = self.evaluate([norm.prob(x), sasnorm.prob(x)])
self.assertAllClose(norm_pdf, sasnorm_pdf)
norm_samps, sasnorm_samps = self.evaluate([
norm.sample(10000, seed=tfp_test_util.test_seed()),
sasnorm.sample(10000, seed=tfp_test_util.test_seed())
])
self.assertAllClose(loc, sasnorm_samps.mean(axis=0), atol=0.1)
self.assertAllClose(norm_samps.mean(axis=0),
sasnorm_samps.mean(axis=0),
atol=0.1)
self.assertAllClose(norm_samps.std(axis=0),
sasnorm_samps.std(axis=0),
atol=0.1)
def test_consistency_random_parameters_no_batch_dims(self):
d = 3
scale_diag = rng.rand(d)
scale_identity_multiplier = np.float64(1.1)
sasnorm = tfd.VectorSinhArcsinhDiag(
scale_diag=scale_diag,
scale_identity_multiplier=scale_identity_multiplier,
skewness=rng.randn(d) * 0.5,
tailweight=rng.rand(d) + 0.7,
validate_args=True)
self.run_test_sample_consistent_log_prob(self.evaluate,
sasnorm,
radius=1.0,
center=0.,
rtol=0.1)
self.run_test_sample_consistent_log_prob(self.evaluate,
sasnorm,
radius=1.0,
center=-0.15,
rtol=0.1)
self.run_test_sample_consistent_log_prob(self.evaluate,
sasnorm,
radius=1.0,
center=0.15,
rtol=0.1)
def test_positive_skewness_moves_mean_to_the_right(self):
d = 10
scale_diag = rng.rand(d) + np.float64(1.0)
loc = rng.randn(d)
sasnorm = tfd.VectorSinhArcsinhDiag(loc=loc,
scale_diag=scale_diag,
skewness=3.0,
validate_args=True)
sasnorm_samps = self.evaluate(
sasnorm.sample(10000, seed=tfp_test_util.test_seed()))
np.testing.assert_array_less(loc, sasnorm_samps.mean(axis=0))
def test_tailweight_small_gives_fewer_outliers_than_normal(self):
d = 10
scale_diag = rng.rand(d)
scale_identity_multiplier = np.float64(0.9)
loc = rng.randn(d)
norm = tfd.MultivariateNormalDiag(
loc=loc,
scale_diag=scale_diag,
scale_identity_multiplier=scale_identity_multiplier,
validate_args=True)
sasnorm = tfd.VectorSinhArcsinhDiag(
loc=loc,
scale_diag=scale_diag,
scale_identity_multiplier=scale_identity_multiplier,
tailweight=0.1,
validate_args=True)
# sasnorm.pdf(x) is smaller on outliers (+-10 are outliers)
x = np.float64([[-10] * d, [10] * d]) # Shape [2, 10]
norm_lp, sasnorm_lp = self.evaluate(
[norm.log_prob(x), sasnorm.log_prob(x)])
np.testing.assert_array_less(sasnorm_lp, norm_lp)
# 0.1% quantile and 99.9% quantile are outliers, and should be more
# extreme in the normal. The 97.772% quantiles should be the same.
norm_samps, sasnorm_samps = self.evaluate([
norm.sample(int(5e5),
seed=tfp_test_util.test_seed(hardcoded_seed=1)),
sasnorm.sample(int(5e5),
seed=tfp_test_util.test_seed(hardcoded_seed=1))
])
np.testing.assert_array_less(np.percentile(norm_samps, 0.1, axis=0),
np.percentile(sasnorm_samps, 0.1, axis=0))
np.testing.assert_array_less(
np.percentile(sasnorm_samps, 99.9, axis=0),
np.percentile(norm_samps, 99.9, axis=0))
# 100. * sp.stats.norm.cdf(2.)
q = 100 * 0.97724986805182079
self.assertAllClose(np.percentile(sasnorm_samps, q, axis=0),
np.percentile(norm_samps, q, axis=0),
rtol=0.03)
self.assertAllClose(np.percentile(sasnorm_samps, 100 - q, axis=0),
np.percentile(norm_samps, 100 - q, axis=0),
rtol=0.03)
def test_consistency_random_parameters_with_batch_dim(self):
b, d = 5, 2
scale_diag = rng.rand(b, d) + np.float64(1.1)
sasnorm = tfd.VectorSinhArcsinhDiag(scale_diag=scale_diag,
skewness=rng.randn(d) * 0.5,
tailweight=rng.rand(b, d) + 0.7,
validate_args=True)
self.run_test_sample_consistent_log_prob(self.evaluate,
sasnorm,
radius=1.0,
center=0.,
rtol=0.1)
self.run_test_sample_consistent_log_prob(self.evaluate,
sasnorm,
radius=1.0,
center=-0.15,
rtol=0.1)
self.run_test_sample_consistent_log_prob(self.evaluate,
sasnorm,
radius=1.0,
center=0.15,
rtol=0.1)
